4 ms·
For some reason we can't "spread the load" of economic activity so that we are not all crammed into dots on the map. I suspect this is partially due to the spe
by tabtab 7y ago
For some reason we can't "spread the load" of economic activity so that we are not all crammed into dots on the map.
I suspect this is partially due to the specialization needed to globally compete. You have to hire lots of specialists to compete and put them in the same building so they are collaborating efficiently. Tele-work hasn't yet lived up to the promise, perhaps because office politics matter, and tele-work hides one from the real going-ons.
- ghaff 7y agoIt really depends. I mostly work remote--technically I'm assigned to an office but don't even have a desk there--but am in a very distributed group within a relatively distributed company. Most of the external people we work with (and our customers) are very distributed anyway. But, yes, on-site is still the default for both good and bad reasons.
- tabtab 7y agoI can see that working once everyone knows the systems and each other's tendencies and habits, but a new person pretty much needs to be embedded to absorb unwritten information.
- nabla9 7y agoYou should look up the term "agglomeration effects" or "agglomeration economiies" This is the stuff that urban economists study and very interesting subject. Cities are like hivemind. The mind works better when things bump together constantly. Network theory can be used to explain some of that. There is more edges between nodes and more integration. Some of the effects is just accidental information transfer when people are more likely to meet by accident. Other are knowledge spillovers like intellectual connections among inventors and diffusing useful knowledge. There are returns to scale, clustering effects. Labor market pooling is important. The easiest to understand is reducing the costs of moving goods over space (Pizza delivery has higher productivity when distances are shorter).